Winter World: The Ingenuity of Animal Survival - Bernd Heinrich (2003)
HIBERNATING SQUIRRELS (HEATING UP TO DREAM)
On several winter mornings in 2002 at dawn, when I sat at my desk looking out into the woods, I saw three gray squirrels emerge, one after the other, from the same leafy nest high in a pine tree. Within a few minutes the trio left unhurriedly, traveling on the bare winter branches of the maples. Like tightrope walkers they balanced on the slim branches and then acrobatically jumped from one tree to the next. They fed on the buds of broad-leafed trees after snipping off the terminal twigs (which they dropped), on sunflower seeds when they were available in the bird feeder, and on acorns on the trees or on the ground. In the spring, a red squirrel has a nest with young in one of my birdhouses. And these are not my only squirrel neighbors.
Within less than a hundred yards from my home in Vermont, and within a mile of my cabin in the Maine woods, live two more species of squirrels, in addition to the flying, gray, and red squirrels. All descended from a common ancestor more than 60 million years ago. They diverged and specialized on different kinds of food. They are what they eat and they hibernate depending on what they eat.
The most common, conspicuous, and noisy of the local squirrels is the little red Tamiasciurus hudsonicus, also called pine squirrel in parts of its range. It is the “sentinel of the taiga” as William O. Pruitt Jr. calls it in a little book titled Animals of the North that I have long treasured. It leaves signs of its presence everywhere: cone bracts of pine and spruce freshly strewn over the surface of the snow, cone cores discarded on a log where tunnels enter the base of an old pine stump. Almost every fresh snowfall is quickly followed by a new sign, and the perpetrator of that sign will likely be perched on a branch next to a trunk above your head. The cheeky little chicoree (still another name for T. hudsonicus) will let loose with a loud sputtering chatter or a churrrrrr that resounds throughout the forest. This will usually be sequenced to a long series of harsh staccato chatter, accompanied by flicks of its fluffy tail over its head and thumping with its hind feet for emphasis. Red squirrels are emphatically active at any month of winter. They appear not to hibernate at all. However, during periods of extreme cold, the woods are silent, and they hole up for days at a time in their subterranean burrows under a stump or tree roots.
Once underground they are almost fully protected from the cold. At least some of their populations, particularly out West, make large caches of seed cones, and with that food they can presumably continue to be active. Yet, cone crops are not reliable every year, and in Maine there are many years, such as the winter of 2001-2002, when I’ve found no caches at all. At these times they feed on the buds of spruce and fir (see Chapter 3). Although there is no guarantee that they do not become inactive with lowered body temperature for a few days, if need be, their emphasis is on fighting the cold by storing food if they can, finding alternate food if they have to, seeking shelter, and growing a thick, rich-rust-colored, insulating fur coat in winter.
The red squirrels’ temporary retreats into tunnels and dens in the winter at subzero temperatures contrasts with the behavior of the other four local squirrel species. Of these, the larger-bodied gray squirrel (Sciurus carolinensis) and the much smaller northern flying squirrel (Glaucomys sabrinus) stay above ground the whole time. Two others, the eastern chipmunk (Tamias striatus) and its very much larger cousin the woodchuck or groundhog (Marmota monax) absent themselves from the cold snowy world above ground for weeks, months, and even to half the year. In general, most ground squirrels hibernate all or most of the winter, whereas tree squirrels, which can still find food on trees, don’t. The stark differences in overwintering biology within this one group of related animals shows that hibernation is less a strategy of avoiding the cold than of what they eat, of weathering famine.
The local squirrel showing the least tendency to hibernate is the now-often suburban gray squirrel. It is active through all months of winter. In the absence of the largess of sunflower seeds at bird feeders, these squirrels will dig through shallow snow to recover acorns, nuts, and maple seeds stored in the fall. If seed crops fail them, they then feed on tree buds and sometimes bark. Food storage, snug leafy well-insulated nests, and large body size give them enough energy resources and means to conserve body heat so that they do not need to hibernate.
Chipmunks are “true hibernators.” Like other ground squirrels, they spend most or all of the winter in a subterranean nest where they curl up, cool down, and become torpid. However, they are not torpid all winter long. If they were, then they would not need to lay up food stores to fuel body heating. Torpid animals don’t eat. The chipmunk’s large cheek pouches indicate an ancient evolutionary commitment to storing food. I do not know how many seeds a chipmunk usually packs into each of its two pouches—I easily inserted sixty black sunflower seeds through the mouth into just one pouch of a roadkill. Chipmunks seldom fail to fill both pouches on any visit to my bird feeder, whereas the gray and red squirrels never carry away even a single seed. Anything they eat, they have to eat in place.
By late fall after a good sugar maple seed year, or after finding a well-stocked bird feeder, chipmunks take trip after trip fully loaded, and all trips lead into the hibernation burrow system that has special granary chambers. These food stores are especially needed in March when the snow is generally still deep. It is then the mating season and the male chipmunks burrow to the surface. There is no new food yet above-snow, but traveling on the crust is easy and those little ground squirrels with the most stockpiled food from the fall can then be the most single-minded in their pursuits.
Normally I don’t see a single chipmunk all winter long. They stay underground, entering into periods of torpor. But torpor is an option, not a necessity or a rule, as was apparent in the winter of 2000-2001 when throughout Maine and Vermont we had an exceptionally large mast crop in the fall. The sugar maples, red oaks, and beech all simultaneously produced bumper seed crops, whereas in many years they produced no seed at all. It was also a winter of exceptionally deep snow. Yet, despite the frequent storms that winter, the chipmunks came to our feeder all winter long.
A chipmunk’s availability of stored food affects whether it remains fully active or enters full torpor (Panuska 1959). But entry into torpor also requires a cold stimulus. Chipmunks are light sleepers; handled torpid chipmunks invariably become roused (Newman 1967). When that happens, their metabolic rates increase as much as fiftyfold within an hour. In contrast to the torpor induced by starvation, as in a nonhibernator close to death, the hibernating chipmunks’ low body temperature is not passive. At an air temperature of 0°C they regulate their body temperature near 6°C, rather than at 37°C when they are active. At air temperatures above 15°C, however, body temperature of hibernating chipmunks is no longer regulated, passively increasing with increasing air temperature (Newman 1967).
Like chipmunks, northern flying squirrels as already mentioned also do not fatten up for winter, nor do they put on a thick insulating fur as red squirrels do. Nor do they lay up stores of food. Instead, they solve the energy problem by huddling in groups in snug nests. Even at -5°C outside the nest, the temperature within the nest is not yet low enough for them to have to shiver to keep warm.
Unlike the eastern chipmunk, some ground squirrels from the western American mountains and deserts enter into hibernation torpor not in response to cold. They begin to hibernate in the hottest, driest part of the year and then continue to stay torpid through the winter (Cade 1963). These squirrels enter hibernation regardless of temperature, and also regardless of the absence or presence of food and water. One of these species, the golden-mantled ground squirrels (Citellus lateralis) has gained fame for the revelations that have emerged for the timing of its hibernation, through the experimental work of Kenneth C. Fisher and his student Eric T. Pengelley. Their subject animal, unlike eastern chipmunks, does not store food but instead fattens up prior to hibernation. So much to eat, so little time. How do they know when to begin? They consult an internal calendar.
Calendar-type timing was suspected when Fisher and Pengelley noted that their squirrels kept under constant light and temperature conditions in the lab at the University of Toronto stopped eating and drinking and went into hibernation in October, at the same time that those outside exposed to the natural environment did. For four consecutive years in one experiment, the squirrels’ cycles of feeding, fattening, and hibernation torpor all continued in the absence of all external cues, in somewhat shortened annual cycle of 324 to 329 days rather than 365 days. The cycling of behavior and physiology occurred in squirrels held at year-round constant temperatures of both 0°C as well as at 22°C in the lab. At a constant 30°C they no longer hibernated, although they still maintained their annual cycle of feeding and fattening that is normally associated with hibernation. These dramatic results showed that the squirrels consult an internal timer, and in analogy with the previously known circadian or daily rhythms, Pangelley and Fisher coined the word “circannual” (circa = approximately, annum = year). Such circannual calendars have since been demonstrated in the timing of bird migration and in other subterranean rodent hibernators. They play a role not only in preparation for deep hibernation but also in arousal from it. The groundhog or woodchuck is an especially well known example.
The groundhog is a large ground squirrel that according to the legend popularized by or of the old Pennsylvania Dutch settlers has an amazingly precise internal calendar. At least one of them, Punxsutawney Phil, emerges punctually from his burrow every year on the second day of February (in Pennsylvania) (at the precise time the news media start to roll TV cameras) to check if he can see his shadow to decide whether or not to go down and sleep for another two weeks.
A groundhog’s survival does indeed depend on accurate scheduling: The squirrel must synchronize its life with the availability of veggies. If possible, it feeds on lettuce, carrots, peas, beans, and other freshly picked produce. Its natural food, grass and weeds, will do only if it can’t get into a garden. In either case, food is available for only about a third of the year. Furthermore, greens don’t store well in the constant moisture under snow and in underground burrows. They’d quickly become a moldy mess. (A relative of rabbits, the pica that lives in mountain areas where there is more wind and dryness has hit on a solution. It gathers greens and dries them to make hay, which is stored and later eaten throughout the winter.) The groundhog has a different solution. It converts the summer greens not to hay, but thanks to a prodigious and adequately preprogrammed appetite, to thick rolls of body fat.
Obesity has its advantages, such as when the animal can be safely inactive in its den. For the rest of the time obesity makes the animal a considerably more attractive meal to predators, all the while compromising its speed and agility. To minimize its duration of obesity, the groundhog must maximize the speed and extent of becoming obese. To be successful in this endeavor, it delays fattening until near the end of the summer. So, it must not only know what to eat, it must also consult a calendar as to when to start eating as if life depended on it. Thus, as in the golden-mantled ground squirrel, a circannual clock is vital for its winter survival.
FOR PROBABLY THE MOST remarkable story of hibernation and winter survival of any mammal, I turn now to the arctic ground squirrel (Spermophilus parryii). This tawny-gold and gray ground squirrel with small white spots is larger than a chipmunk and smaller than a woodchuck. It is the northernmost mammalian hibernator across the Canadian and Siberian tundra. For eight months of the year this squirrel curls up into a ball close to the ice of the permafrost, and maintains a body temperature at or below the freezing point of water. Brian M. Barnes and colleagues at the University of Alaska at Fairbanks have for many years tried to decipher how these animals survive. They have studied them in the field at the Toolik Lake station at the foothills of the Brooks Range, and in enclosures and in the lab at Fairbanks.
Like other ground squirrels, this species digs hibernation burrows and builds underground nests. However, because of the permafrost of their environment, the squirrels cannot dig deep enough to escape the subzero temperatures of winter. Instead, in late summer and autumn, when the temperatures are merely freezing, they dig down into the soil. The temperature surrounding them declines through fall and winter and continuous records of body temperature in squirrels equipped with radio transmitters indicate that body temperature of torpid squirrels declines in parallel with the sinking soil temperature. Remarkably, however, as burrow temperatures continue to decline to -15°C by December, the squirrel’s body temperature does not continue to decline, to -15°C. Instead, it stabilizes at between -2° and 2.9°C. That is, the torpid squirrel’s body temperature is then no longer allowed to be passive. It is regulated some eight to nine degrees lower than a hibernating chipmunk’s but twelve to thirteen degrees above soil temperature. No other animal had previously been shown to regulate its body temperature near 0°C, much less two or more degrees below the freezing point of water. Furthermore, the squirrels did not turn into blocks of solid ice as their core body temperatures declined to as much as -2.9°C. Barnes wondered if they might have antifreeze. To find out, he removed blood plasma from hibernating squirrels and tested its freezing point in the lab. The blood turned to ice at approximately -0.6°C. It therefore did not contain antifreeze. These results deepen the mystery of winter survival: Why should the blood freeze in the lab but not in the animal? The riddle is not yet solved, but the best tentative explanation so far is that the squirrels supercool.
Pure water has a freezing point of 0°C (32°F). Adding one mole (molecular weight) of a substance to a liter of water lowers its freezing point by -1.86°C. Although pure water and solutions of specific concentration have precisely predictable freezing (and melting) points, it is sometimes possible to lower temperatures below the predicted freezing point without having ice crystals form. Such solutions are said to be “supercooled.” Supercooling occurs due to the absence of “nucleation sites”—places where ice crystals can begin to grow. The best nucleation sites are other ice crystals. Thus, if one adds an ice crystal, say a snowflake, to a vial of pure liquid water that is supercooled to -10°C, then the whole vial full of water will turn instantly into a solid block of ice. But this ice won’t melt until it is heated to 0°C. This difference between freezing and melting points (called thermal hysteresis) defines supercooling. Supercooled liquids are unstable—they can turn to ice unpredictably and with little apparent provocation. Mere stirring can be enough. The greater the thermal hysteresis, the greater likelihood that freezing will occur, and the quicker the sample “flashes” into ice, giving off a measurable pulse of heat in the process as the energy of the motion of the liquid molecules is released when they stop their motion after release from the ice crystal lattice.
The absence of antifreeze in the blood of the squirrels, and hence the likelihood of supercooling to as much as 1° to 2°C ought to be risky. A single stray ice crystal in the blood could mean death. Why do squirrels risk it? Why don’t they regulate their body temperature 1° to 2°C higher, to avoid supercooling and thus be immune to turning into a block of ice and being killed? Barnes believes the advantage that outweighs the cost is related to energy economy; supercooling to -2°C would save the squirrels ten times the energy expended by maintaining a body temperature of 0°C (Barnes 1989). The squirrels also have a mechanism that reduces the risks normally associated with supercooling. Generally freezing would start at the coldest point, such as a toe, and Barnes has nucleated (started the freezing process) squirrels’ toes and found that the animals were then alerted—they rewarmed quickly before the ice could spread.
Barnes’s other remarkable discovery was that the squirrels can and do arouse spontaneously to warm themselves up from a body temperature of less than 0°C, to heat themselves all the way up to their body temperature when active, 37°C. Many other animals can survive cooling to or below 0°C, but none had ever been able to spontaneously arouse unless they were first artificially warmed by being taken to much higher air (and body) temperatures, where the shivering response becomes possible.
Although energy economy helps to explain the squirrel’s low body temperature, far from behaving in a manner strictly in accord with just saving energy, the arctic ground squirrels appeared to squander energy by warming up to 37°C from subzero temperatures about a dozen times throughout their winter hibernation. Each time they spent about a day being fully warmed and required another to cool back down. These periodic warmings are calculated to cost the animals over half of the fat reserves they had built up during the summer. Why do they bother? It had previously been shown that this behavior is not unique to arctic ground squirrels. Indeed, no mammal hibernator avoids such periodic bouts of normal body temperature during a winter’s hibernation. Therefore, given the high energetic costs of warming up and staying warm for a day or so, the behavior seemed most peculiar. It must buy something precious. And this is where Barnes and his colleagues’ fourth remarkable discovery comes in. Although still controversial, the hypothesis is that the animals warm up to sleep!
Since the early 1950s two different kinds of sleep have been defined. One is “rapid eye movement” (REM) sleep, also called “dreaming sleep” or “deep sleep.” The other is defined as “nonrapid eye movement” (NREM) sleep, or “light” or “ordinary sleep.” These two types of sleep have been studied in humans by taking voltage measurements from the scalp surface. Brain electrical activity records, called electroencephalograms (EEGS), are plotted against time. In humans, apes, and cats, where the EEG brain-wave patterns have been most studied, there is a progression of very different patterns from awake to four arbitrary stages presumed to represent depth of sleep, with the awake state showing very low voltage amplitude and high frequency (8-13 waves per second) and deep sleep showing a high-voltage amplitude, low-frequency wave pattern. However, when the EEG sleep patterns are followed throughout the night, periods occur in which low amplitude brain-wave patterns reappear that resemble those during wakefulness. It is during these, the REM sleep periods, that the rapid eye movements, and often increased heart rate, changes the breathing pattern, and muscle twitching occurs. The animal is then dreaming.
Animals in hibernation torpor do not show the EEG sleep patterns. Instead, as their body temperature drops and they enter torpor, there is a gradual reduction of voltage until brain electrical activity eventually disappears, as if they were dead. However, although there is no spontaneous brain activity (Lyman and Chatfield 1953), the animals must still be able to generate at least some electrical activity in the nervous system, or else they could never arouse.
Teaming up with neurobiologists H. Craig Heller and Serge Daan, Barnes recorded electroencephalograms of squirrels going into and coming out of hibernation. Squirrels entering hibernation showed typical sleep patterns, and then as they cooled down their brain waves disappeared and their EEGs then resembled those in humans who would be considered brain dead. However, once rewarmed by shivering, after having been in hibernation torpor for about a month, the squirrels spent most of the day showing the brain patterns associated with rapid eye movements (REM) during dreaming in humans. Had they heated up to sleep, or to dream? If so, why do they sleep or dream? Why do we? It’s one of the large remaining biological mysteries that probably relates in some way to how the brain works to consolidate, edit, delete, and store memory.
It seems ironic that a hibernator has evolved the astounding capacity to arouse from subzero temperatures normally only encountered in winter in order to sleep. If the animals did not need to arouse they could stay torpid until spring and save much energy. Since they don’t stay continually torpid despite the obvious energy economy they would experience by doing so, there is apparently some great cost to long torpor or sleep deprivation.
The hibernating arctic ground squirrels may hold keys to the riddle of why we need sleep, and also some medical problems, such as stroke. In hibernating ground squirrels, it is difficult to detect any heartbeat. It is difficult to tell if the animal, with a body temperature less than the freezing point of water, is dead or alive. In a deathlike state of torpor, the animals are cold little balls in which there is only a minute trickle of blood to the brain. In humans when a blood clot or a ruptured blood vessel interrupts blood flow to a part of the brain, there is an almost immediate die-off of brain cells, because our brain cells require a continuous supply of oxygen and glucose that the continuously flowing blood normally supplies. Hypoxia (insufficient oxygen) is the primary deleterious consequence of a stroke, but not in a hibernating squirrel. In the hibernating squirrel’s brain there is a metabolic shutdown, so that lack of oxygen and nutrient is less harmful. Do they warm up to oxygenate the brain?
The primary metabolic shutdown in hibernators is due simply to the temperature drop. Human car crash victims who fall unconscious through the ice and whose brain is immediately chilled are also able to survive prolonged hypoxia. But there are also active metabolic processes in the squirrels; brain tissues show suppressed protein synthesis even when warmed to 37°C. A second recent interesting finding is that hibernating squirrels as well as hibernating turtles accumulate five times as much ascorbic acid (vitamin C) in their, as opposed to human, brains. When the squirrels arise from torpor, the vitamin C levels return to normal within hours. It is thought that the vitamin C, a powerful antioxidant, protects the brain from the sudden rush of oxygen that they take in after their long oxygen “fast.”
Barnes’s interest in hibernation research is motivated by the discovery of basic phenomena, not practical applications. Many others’ interest in the hibernating ground squirrel is clinically, rather than intellectually, motivated. They wonder how halting blood supply and hence oxygen and glucose to the brain might be relevant to treating stroke victims, where insufficient oxygen to the brain is the main cause of cell death. They wonder how hibernating animals maintain strong bones despite months of inactivity, why their blood clots so slowly, and why they accumulate huge amounts of vitamin C in their brains and cerebrospinal fluids. I suspect that, with biology becoming ever more applied, research relating to such questions of how to harness hibernation will be increasingly funded in preference to those inquiries motivated purely by intellectual curiosity. That is unfortunate and shortsighted.